Self-adaptive parameter and strategy based particle swarm optimization for large-scale feature selection problems with multiple classifiers

Xue, Yu; Tang, Tao; Pang, Wei; Liu, Alex X.

doi:10.1016/j.asoc.2019.106031

Cited by 128 publications

(58 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For instance, in the SpamBase dataset on the KNN classifier, the proposed method obtained a 93.99% classification accuracy. In contrast, for PSO-based [73], ACO-based [75], and ABC-based [78] methods, these values were reported 92.54%, 91.81%, and 90.35%, correspondingly. Moreover, Figs.…”

Section: Resultsmentioning

confidence: 79%

“…For a fair evaluation, all of the methods examined in this section were selected from among wrapper-based methods. These wrapper-based methods include PSO-based [73], ACO-based [75], and ABC-based [78]. These are state-of-the-art EA-based feature selection methods.…”

Section: Resultsmentioning

confidence: 99%

“…In this method, an attempt has been made to use the advantages of coating and filter methods simultaneously. In [73] a PSObased feature selection with multiple classifiers is proposed to improve for increasing the classification accuracy and reducing computational complexity. In this paper, a new Self-Adaptive Parameter and Strategy are used to deal with the issue of feature selection in a high-dimensional dataset.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A novel community detection based genetic algorithm for feature selection

2021

View full text Add to dashboard Cite

The feature selection is an essential data preprocessing stage in data mining. The core principle of feature selection seems to be to pick a subset of possible features by excluding features with almost no predictive information as well as highly associated redundant features. In the past several years, a variety of meta-heuristic methods were introduced to eliminate redundant and irrelevant features as much as possible from high-dimensional datasets. Among the main disadvantages of present meta-heuristic based approaches is that they are often neglecting the correlation between a set of selected features. In this article, for the purpose of feature selection, the authors propose a genetic algorithm based on community detection, which functions in three steps. The feature similarities are calculated in the first step. The features are classified by community detection algorithms into clusters throughout the second step. In the third step, features are picked by a genetic algorithm with a new community-based repair operation. Nine benchmark classification problems were analyzed in terms of the performance of the presented approach. Also, the authors have compared the efficiency of the proposed approach with the findings from four available algorithms for feature selection. Comparing the performance of the proposed method with three new feature selection methods based on PSO, ACO, and ABC algorithms on three classifiers showed that the accuracy of the proposed method is on average 0.52% higher than the PSO, 1.20% higher than ACO, and 1.57 higher than the ABC algorithm.

show abstract

Section: Resultsmentioning

confidence: 79%

Section: Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A novel community detection based genetic algorithm for feature selection

2021

View full text Add to dashboard Cite

show abstract

“…For a fair evaluation, all of the methods examined in this section were selected from among wrapper-based methods. These wrapper-based methods include PSO-based [61], ACO-based [62], and ABC-based [63] 4…”

Section: Experiments and Discussionmentioning

confidence: 99%

A Novel Community Detection Based Genetic Algorithm for Feature Selection

Rostami

Berahmand

Forouzandeh

2020

Preprint

View full text Add to dashboard Cite

The selection of features is an essential data preprocessing stage in data mining. The core principle of feature selection seems to be to pick a subset of possible features by excluding features with almost no predictive information as well as highly associated redundant features. In the past several years, a variety of meta-heuristic methods were introduced to eliminate redundant and irrelevant features as much as possible from high-dimensional datasets. Among the main disadvantages of present meta-heuristic based approaches is that they are often neglecting the correlation between a set of selected features. In this article, for the purpose of feature selection, the authors propose a genetic algorithm based on community detection, which functions in three steps. The feature similarities are calculated in the first step. The features are classified by community detection algorithms into clusters throughout the second step. In the third step, features are picked by a genetic algorithm with a new community-based repair operation. Nine benchmark classification problems were analyzed in terms of the performance of the presented approach. Also, the authors have compared the efficiency of the proposed approach with the findings from four available algorithms for feature selection. The findings indicate that the new approach continuously yields improved classification accuracy.

show abstract

“…Due to its simplicity, the PSO algorithm has attracted the interest of many researchers over the past few decades. As a result, PSO became one of the predominant swarm algorithms applied to various problems including task allocation [3], image processing [4], feature selection [5] and robotic applications [6]. However, the standard PSO algorithm's capability to solve certain kinds of complex problems (e.g.…”

Section: Introductionmentioning

confidence: 99%

HIDMS-PSO: A New Heterogeneous Improved Dynamic Multi-Swarm PSO Algorithm

Varna

Husbands

2020

2020 IEEE Symposium Series on Computational Intelligence (SSCI)

View full text Add to dashboard Cite

In this paper, a variant of the particle swarm optimisation (PSO) algorithm is introduced with heterogeneous behaviour and a new dynamic multi-swarm topological structure. The new topological structure enables the algorithm to have more control over the interaction and information exchange between the particles to reduce the loss of diversity and avoid premature convergence. In the new algorithm, the population is initially divided into two sub-populations, first sub-population is further divided into sub-swarms that are formed using the introduced topological structure. The particles of sub-swarms are guided using heterogeneous behaviour by selecting various exemplars. The second sub-population employs the classical PSO search with local and global information to simulate a homogenous behaviour. There is information flow between the two subpopulations. The algorithm was tested on the CEC2005 and CEC2017 test suites with comparison against various state-ofthe-art PSO variants and other state-of-the-art meta-heuristics. The experimental results show that for the two test suites, the proposed algorithm outperformed the majority of the state-ofthe-art algorithms on most problems.

show abstract

Self-adaptive parameter and strategy based particle swarm optimization for large-scale feature selection problems with multiple classifiers

Cited by 128 publications

References 60 publications

A novel community detection based genetic algorithm for feature selection

A novel community detection based genetic algorithm for feature selection

A Novel Community Detection Based Genetic Algorithm for Feature Selection

HIDMS-PSO: A New Heterogeneous Improved Dynamic Multi-Swarm PSO Algorithm

Contact Info

Product

Resources

About